Reassessing Inflation Targets: Welfare Trade-offs and Evidence from Ghana

This paper studies how a numerical inflation target should be reassessed when an inflation-targeting regime remains useful but the welfare-maximizing target is uncertain. Ghana provides the quantitative application. I first estimate a structural VAR and identify improvements in the effectiveness of the nominal anchor using sign and zero restrictions. The unrestricted output response provides little evidence that better anchoring requires a large or persistent contraction in economic activity. I then develop and estimate the Ghana Inflation Target Assessment Model (GITAM), a nonlinear small-open-economy New Keynesian model in which alternative inflation targets generate different steady states and welfare outcomes. The model incorporates domestic- and import-price adjustment costs, nominal wage adjustment, heterogeneous downward nominal wage rigidity, exchange-rate and external-financing channels, and a target-consistent monetary-policy rule. The welfare-maximizing target depends importantly on the strength of downward nominal wage rigidity and on uncertainty in the SMM-estimated propagation parameters. Across 30 labour-market calibrations, the optimal target ranges from zero to 8.5 percent, but 29 imply an optimum below Ghana's current 8 percent midpoint. Under local SMM parameter uncertainty, the median welfare-maximizing target is 4 percent, with a 90 percent range of 2.5 to 8.5 percent, and 92 percent of admissible parameter draws place the optimum below 8 percent. The broader result is that uncertainty about the exact welfare optimum need not prevent informative ranking of an inherited policy target.